INTERNATIONAL JOURNAL OF MOlecular medicine 42: 615-624, 2018

Inhibitory effects of luteolin‑4'‑O‑β‑D‑glucopyranoside on and A2 ‑mediated amplification of activation in vitro

HUANJUN XU1, HONG LU2, XIAOCUI ZHU1, WEI WANG1, ZHOUMIAO ZHANG1, HUIZHENG FU3, SHUANGCHENG MA4, YUEHUA LUO3 and JIANJIANG FU1

1Department of Pharmacology, School of Pharmacy; 2Network and Educational Technology Center, Jiangxi University of Traditional Chinese Medicine, Nanchang, Jiangxi 330004; 3Jiangxi Provincial Institute for Drug Control, Nanchang, Jiangxi 330029; 4National Institutes for Food and Drug Control, Beijing 100050, P.R. China

Received November 1, 2017; Accepted March 27, 2018

DOI: 10.3892/ijmm.2018.3634

Abstract. Platelet activation and subsequent accumulation may be associated with its dual‑receptor inhibition on P2Y12 at sites of vascular injury are central to thrombus formation, and TP receptors. which is considered to be a trigger of several cardiovascular diseases. Callicarpa nudiflora (C. nudiflora) Hook is a Introduction traditional Chinese medicinal herb for promoting blood circulation by removing blood stasis. In our previous study, are small, a nucleate blood cells, the major role of several compounds extracted from this herb, including which is in and owing to their capacity luteolin‑4'‑O‑β‑D‑glucopyranoside (LGP), were revealed to adhere to damaged blood vessels and to accumulate at sites to exert inhibitory effects on adenosine diphosphate of injury (1). However, platelets are also important contributors (ADP)‑induced platelet aggregation. The aim of present study to thrombotic disorders, including atherothrombosis, which was to confirm these antiplatelet effects and elucidate the are the final events complicating cardiovascular diseases (2‑4). potential mechanisms. Using a platelet‑aggregation assay, it Upon vascular injury, platelets are exposed to the subendothe- was revealed that LGP significantly inhibited platelet aggre- lium, and several , including adenosine diphosphate gation induced by ADP, and . It was (ADP) and thrombin, are generated at the injury site, which also found that LGP exhibited marked inhibitory effects on can stimulate platelet adhesion, activation and aggregation. the activation of αIIbβ3 integrin, the secretion of serotonin from Adherent, activated platelets recruit additional platelets to the granules, and the synthesis of . In addition, growing thrombus (5,6). The uncontrolled progression of these the results showed that LGP suppressed Ras homolog family processes through a series of self‑sustaining amplification member A and phosphoinositide 3‑kinase/Akt/glycogen loops can initiate unrestrained platelet activation and aggrega- synthase kinase 3β signal transduction. Data from a radio- tion, and eventually lead to thromboembolic events (7,8). It has labeled ‑binding assay indicated that LGP exhibited been demonstrated clinically that the use of antiplatelet agents apparent competing effects on thromboxane receptor (TP) to prevent and/or reverse platelet aggregation is a successful and P2Y12 receptors. In conclusion, the data presented here strategy for the prevention of thrombosis (7,8). However, demonstrated that LGP, a natural compound from C. nudiflora due to their disturbance of the thromboregulatory balance, Hook, inhibited the development of platelet aggregation and existing antiplatelet drugs can cause severe side effects, the amplification of platelet activation. These inhibitory effects majority of which limit the efficacy and safetyof these drugs. Uncontrolled hemorrhage is the most frequent side effect of /antiplatelets (9,10). Therefore, understanding the molecular mechanisms of platelet activation and identi- fying novel techniques for platelet inhibition remain critically important. Correspondence to: Dr Jianjiang Fu, Department of Pharmacology, Natural products remain a major resource and have become School of Pharmacy, Jiangxi University of Traditional Chinese Medicine, 818 Meiling Road, Nanchang, Jiangxi 330004, P.R. China increasingly important for novel drug identification. The E‑mail: [email protected] Callicarpa genus includes ~190 extant species (11). Among these, there are >10 medical herbs, and the majority of these

Key words: platelet activation, P2Y12 receptor, thromboxane A2 have hemostasis‑associated usage (11). Callicarpa nudiflora receptor, Callicarpa nudiflora, luteolin‑4'‑O‑β‑D‑glucopyranoside (C. nudiflora) Hook, which is one of the medical herbs of Callicarpa with a long history, is used for eliminating stasis in order to subdue swelling and hemostasis (11). According 616 XU et al: ANTIPLATELET EFFECTS OF LGP to traditional Chinese medicine theory, eliminating stasis to subdue swelling is similar to the effect. Led by these concepts, the present study hypothesized that antiplatelet activities may contribute to the traditional usage of this plant. In our previous study, hundreds of constituents were screened, including several derivatives of luteolin. It was found that two novel triterpenoids, extracted from the leaves of C. nudiflora, showed inhibitory effects on ADP‑induced platelet activation (12,13). In the present study, it was shown that one of the derivatives of luteolin, 1,6‑di‑O‑caffeoyl‑β‑D‑glucopyranoside (LGP) exhibited potent inhibitory effects on platelet activation, and it was Figure 1. Molecular structure of luteolin‑4'‑O‑β‑D‑glucopyranoside. demonstrated that the effects of this natural compound may be mediated by dual receptor antagonism on P2Y12 receptor and thromboxane A2 (TXA2) receptor (TP). The PRP was adjusted to a platelet count of 400x109 platelets/l Materials and methods by diluting in PPP.

Drugs and chemicals. Luteolin‑4'‑O‑β‑D‑glucopyranoside Platelet aggregation. Platelet aggregation in 96‑well plates was (LGP) was isolated from the leaves of C. nudiflora Hook measured using a modified light transmission method (14,15). and its molecular structure is shown in Fig. 1. The purity Briefly, the PRP (400x109 platelets/l) was incubated with 40, of LGP was ≥95%, as determined by high‑performance 80 and 160 µM of LGP, antagonists (positive controls) or liquid chromatography. A stock solution was prepared by dissolvent for 15 min at 37˚C, respectively. The optical density dissolving the LGP in 100% dimethyl sulfoxide (DMSO) (OD) was then determined at 595 nm and marked as OD1. and was used throughout the investigation. The final DMSO Platelet aggregation was induced by the following agonists: concentration did not exceed 0.1%. ADP, arachidonic acid ADP (10 µM), U46619 (1 µM) or AA (600 µM), and OD (OD2) (AA), (an antagonist of the P2Y12 receptor) and was determined again at 595 nm every 30 sec for 15 min, BM‑531 (an antagonist of the TP receptor), were purchased with 15 sec incubation and 15 sec shaking between readings. from Sigma‑Aldrich; EMD Millipore (Billerica, MA, USA). In addition, the OD value at 595 nm was determined for the

U46619 was the product of Tocris Bioscience (Bristol, UK). same volume of PPP, and was marked as OD3. All experi- [3H]‑2‑methylthioadenosine diphosphate ([3H]‑2‑MeS‑ADP) ments were performed at least three times. The percentage and [3H] SQ‑29548 were purchased from GE Healthcare Life of aggregation was calculated using the absorbance of PRP Sciences (Chalfont, UK) and PerkinElmer, Inc. (Waltham, without as 0% aggregation and the absorbance of PPP MA, USA), respectively. as 100% aggregation. The relative aggregation was expressed

using the following formula: Relative aggregation (%)=[(OD1‑ Animals. A total of 80 female Sprague‑Dawley rats (aged OD2)/(OD1‑OD3)] x100. 6‑8 weeks old and weighing 180‑220 g) were obtained from

Vital River Laboratories (Beijing, China) and maintained Activated αIIbβ3 integrin abundance. αIIbβ3 integrin is under pathogen‑free conditions in the Animal Center of Jiangxi expressed on the surface of platelets, which reflects platelet University of Traditional Chinese Medicine (Nanchang, activation or secretion from platelet granules. This was China). All the animals were maintained in a 12 h light/dark determined by the measurement of fluorescent agent‑labeled cycle at room temperature (25±2˚C) in 60% humidity. The antibody binding, as described previously (16,17). For animals were allowed water ad libitum and were fed a standard measurement of the expression of αIIbβ3 integrin, briefly, laboratory diet. The protocol used for animal experiments 50 µl activated platelets (400x109 platelets/l) were pretreated (JZAEC‑2016‑0031) was approved by the Animal Ethics for 15 min with DMSO or the indicated concentrations (20, Committees of Jiangxi University of Traditional Chinese 40, 80 and 160 µM) of LGP, and then fixed for 30 min in Medicine, and all animal experiments were performed in strict 0.5% paraformaldehyde. Following washing once in incu- accordance with the requirements of this protocol. bation buffer, the fixed platelets were added into 96‑well plates, and incubated with Oregon Green‑labeled fibrinogen Preparation of rat platelets. Blood was collected from the (Molecular Probes; Thermo Fisher Scientific, Inc., Waltham, abdominal aorta of anesthetized rats into a vacuum blood MA, USA) for 15 min at room temperature without shaking. collection tube, which allowed 10% blood volume with 3.8% The total fluorescence was determined by measuring the fluo- sodium citrate as . The citrated blood was then rescence of the plate on a multi‑label counter (VICTOR3™; centrifuged (Allegra™ X‑12R centrifuge; Beckman Coulter, PerkinElmer, Inc.). Subsequently, the plates were centrifuged Inc., Brea, CA, USA) at 110 x g for 15 min at 4̊C to obtain at 1,000 x g for 10 min at 4˚C to remove platelets from the platelet‑rich plasma (PRP), and the quantity of platelets in the supernatant. The supernatants were transferred to separate PRP was determined using the automatic blood cell analyzer plates and the fluorescence was determined (background or (HEMAVET 950FS; Drew Scientific, Miami Lakes, FL, nonspecific). Specific Oregon Green binding was determined USA). Platelet‑poor plasma (PPP) was obtained by a second by subtracting the background fluorescence from the total centrifugation of the remaining blood (1,000 x g, 10 min, 4˚C). fluorescence. INTERNATIONAL JOURNAL OF MOlecular medicine 42: 615-624, 2018 617

Measurement of serotonin (5‑HT) secretion. 5‑HT secre- (ARHGEF1) antibody (1:1,000; Abcam, cat. no. ab220892), tion was determined using a Serotonin ELISA kit (Abcam, anti‑Ras homolog family member A (RhoA) antibody (1:1,000; Cambridge, MA, USA; cat. no. ab133053) according to the Abcam; cat. no. ab54835), anti‑RhoA antibody (1:1,000; manufacturer's protocol. The assay procedure is based on phospho S188; Abcam; cat. no. ab41435), anti‑Rho‑associated the competition between an alkaline phosphatase‑conjugated kinase 1 (ROCK1) antibody (1:1,000; Abcam; cat. no. ab45171), 5‑HT (supplied) and a non‑labeled antigen (5‑HT extracted anti‑ROCK1 (phospho T455+S456) antibody (1:1,000; Abcam; from PRP) for a fixed number of antibody binding sites on cat. no. ab203273), anti‑phosphoinositide 3‑kinase (PI3K) p85 the micro‑titer plate. First, a curve of the OD405 of 5‑HT, (phospho Y607) antibody (1:1,000; Abcam; cat. no. ab182651), compared with its concentration in the standard wells, was anti‑pan‑AKT antibody (1:1,000; Abcam; cat. no. ab8805), plotted. Subsequently, 50 µl platelets (400x109 platelets/l) were anti‑pan‑AKT (phospho T308) antibody (1:1,000; Abcam; treated with DMSO or the indicated concentrations (20, 40, 80 cat. no. ab38449), anti‑AKT1 (phospho S473) antibody and 160 µM) LGP for 15 min, followed by 3 min incubation (1:1,000; Abcam; cat. no. ab81283) and anti‑glycogen synthase with agonists (10 µM ADP or 1 µM U46619). The reaction kinase 3β (GSK3β) antibody (1:1,000; phospho S9; Abcam; was terminated by snap freezing. Following thawing at room cat. no. ab75814). The secondary antibodies used in the present temperature, the samples were centrifuged at 3,000 x g for study were as follows: Goat polyclonal secondary antibody 10 min at 4˚C. The supernatants were used for the measure- to mouse IgG (1:5,000; Abcam; cat. no. ab6789) and goat ment of 5‑HT release. By comparing the absorbance of the anti‑rabbit IgG (1:5,000; Abcam; cat. no. ab6721). samples with the standard curve, the 5‑HT concentration in the unknown samples was determined, with data representa- Receptor‑binding assay. The effects of LGP on P2Y12 tive of at least five independent experiments. ADP receptor binding were determined by the binding of [3H]‑2‑MeS‑ADP to rat platelets with a filter technique to 3 3 Measurement of TXA2 synthesis. In the present study, TXB2, separate the free from bound [ H]‑2‑MeS‑A DP. [ H]SQ‑29548 the stable metabolite of TXA2, was measured to reflect the level (PerkinElmer, Inc.) was also used in to assess the effects 9 of TXA2. PRP (400x10 platelets/l) was pretreated with DMSO of LGP on TXA2 receptor binding activities. Briefly, PRP or various concentrations of LGP (20, 40, 80 and 160 µM) for (1x109 platelets/ml) was incubated with [3H]SQ‑29548 (40 nM 15 min at 37˚C, and was stimulated with ADP (10 µM) or final concentration) in a total of 400 µl Tyrode's buffer (pH 7.2) U46619 (1 µM) at 37˚C for 3 min whilst stirring. The reaction for 30 min at room temperature. Subsequently, indicated was also terminated by snap freezing. Following thawing at concentrations of LGP were added and incubated for 40 min room temperature and centrifuging at 3,000 x g for 10 min at room temperature to compete binding between agonists and at 4˚C, the supernatants were diluted (1:20) with the assay their receptors. The binding assays were terminated by rapid buffer in the TXB2 ELISA kit (Cayman Chemical Company, filtration on Packard GF‑B filters (Packard Instrument Co., Ann Arbor, MI, USA). TXB2 was measured according to the Inc., Meriden, CT, USA). The filters were then placed in plastic manufacturer's protocol. scintillation vials containing an emulsion‑type scintillation mixture (4 ml) and the radioactivity, representing the binding Western blot analysis. PRP (400x109 platelets/l), pretreated of [3H]SQ‑29548 to TP receptor (B), was detected by Tri‑Carb® with DMSO or various concentrations (40, 80 and 160 µM) Liquid Scintillation (PerkinElmer, Inc.). The radioactivity of LGP for 15 min, were stimulated for 3 min with agonists, and [3H]SQ‑29548 (40 nM final concentration) in DMSO‑treated the reaction was terminated by rapid freezing of the sample in a platelets served as the total binding (Bt) of [3H]SQ‑29548 to dry ice‑ethanol bath. Following thawing at room temperature, the TP receptor. Non‑specific binding (Bns) was defined as the the samples were centrifuged at 3,000 x g for 10 min at 4˚C. total radioactivity measured in the presence of 100 µM (final The platelets were rinsed twice with PBS, and total concentration) unlabeled SQ‑29548. The specific binding rate were extracted with lysis buffer. Aliquots of each platelet lysate (Bs) of [3H]SQ‑29548 to the TP receptor was calculated using containing equal quantities of (ranging between 500 the following formula: Bs = (B ‑ Bns)/Bt x 100. and 750 µg between experiments) were added to SDS‑PAGE For the P2Y12 ADP receptor binding assay, a similar gels (ranging between 8 and 12%), and then transferred onto procedure to the TP receptor binding assay was used. PRP hybond nitroblotting membranes and subjected to western blot (1x109 platelets/ml) was prepared, as previously described, and analysis. Membranes were blocked using 5% non‑fat dried incubated with [3H]‑2‑MeS‑ADP (5 nM final concentration) in milk for 1 h at room temperature and subsequently incubated a total of 400 µl Tyrode's buffer (pH 7.2) for 30 min at room with primary antibodies overnight at 4˚C. Following washing temperature. Subsequently, indicated concentrations of LGP, with 0.5% TBST three times, the membranes were incubated DMSO or 2‑MeS‑ADP (5 µM, final concentration) were added with horseradish peroxidase‑conjugated secondary antibodies for an additional 40 min at room temperature. Following for 2 h at room temperature. The immunoreactive bands filtration on Packard GF‑B filters, radioactivity was detected were detected using an enhanced chemiluminescence kit by Tri‑Carb® Liquid Scintillation (PerkinElmer, Inc.) and the 3 (EMD Millipore). β‑actin (1:1,000; Santa Cruz Biotechnology, specific binding rate of [ H]‑2‑MeS‑ADP to the P2Y12 receptor Inc., Dallas, TX, USA; cat. no. SC‑130656) served as an internal was calculated. control. The signal intensities of the bands of interest were quantified and normalized toβ ‑actin using the Image‑Pro Plus Data presentation and statistical analysis. Data are presented software version 6.0 (Media Cybernetics, Inc., Rockville, MD, as the mean ± standard error of the mean; n represents the USA). The primary antibodies used in the present study were number of independent experiments. Statistical signifi- as follows: Anti‑Rho guanine nucleotide exchange factor 1 cance was determined using one‑way analysis of variance. 618 XU et al: ANTIPLATELET EFFECTS OF LGP

Figure 3. Effects of LGP on ADP‑ and U46619‑mediated αIIbβ3 integrin activation. (A) Effects of LGP on ADP‑induced αIIbβ3 integrin activation. (B) Effects of LGP on U46619‑induced αIIbβ3 integrin activation. Each assay was performed in triplicate. **P<0.01, compared with control; ##P<0.01, compared with ADP or U46619. Vehicle (DMSO) was used as a control, and ticagrelor and BM‑531 served as the positive controls. LGP, luteolin‑4'‑O‑β‑D‑glucopyranoside; ADP, adenosine diphosphate; OG‑FGN, Oregon Green‑labeled fibrinogen.

Figure 2. Effects of LGP on platelet aggregation induced by several agonists. (A) Effects of LGP on ADP‑induced platelet aggregation. (B) Effects of LGP concentration‑dependent inhibition of aggregation induced by on U46619‑induced platelet aggregation. (C) Effects of LGP on AA‑induced U46619 and AA with IC50 values of 61.7±1.2 and 81.7±1.1 µM platelet aggregation. Each assay was performed in triplicate. Vehicle at 540 sec, respectively (Fig. 2B and C). (DMSO) was used as a control, and ticagrelor and BM‑531 served as the positive controls. LGP, luteolin‑4'‑O‑β‑D‑glucopyranoside; ADP, adenosine diphosphate; AA, arachidonic acid. LGP suppresses ADP and U46619‑mediated αIIbβ3 integrin activation in rat platelets. It is widely accepted that integrin

αIIbβ3‑mediated outside‑in signaling is the most important amplifier of platelet activation. To confirm the effects of LGP

Dose‑response curves were generated using GraphPad Prism on outside‑in signal transduction, the active integrin αIIbβ3 software (version 4.0; GraphPad Software, Inc., La Jolla, CA, on the platelet surface was assessed by the measurement of

USA). The IC50 value for each agent was determined from fibrinogen binding. As shown in Fig. 3A, the level of integrin three different concentrations of the agent using Schild anal- αIIbβ3 was negligible at the surface of resting platelets. There ysis using GraphPad Prism software. P<0.05 was considered was a sharp increase in the level of integrin αIIbβ3 following to indicate a statistically significant difference. ADP (10 µM) treatment, and a significant attenuation in the presence of LGP and the positive control ticagrelor. Similarly,

Results the level of integrin αIIbβ3 was significantly increased by treat- ment with U46619 (1 µM). Again, the effect was significantly LGP inhibits ADP‑, U46619‑ and AA‑induced platelet inhibited in the presence of LGP and BM‑531 (Fig. 3B). These aggregation. The initial experiment defined the effects of data indicated that LGP inhibited the activation of integrin

LGP on platelet aggregation induced by several agonists. Rat αIIbβ3 in a concentration‑dependent manner when the platelets platelets were isolated and platelet aggregation was observed. were stimulated by ADP or U46619. These results are compat- The platelets were pretreated with LGP for 15 min and were ible with the results of the aggregation assay. incubated with 10 µM ADP. As shown in Fig. 2A, 10 µM ADP (to PRP) produced typical aggregation curves. Ticagrelor and LGP inhibits5‑HTrelease stimulated by ADP and U46619. LGP significantly inhibited aggregation in a dose‑dependent The platelets pretreated by LGP were incubated with agonists manner, and the IC50 value of LGP at 540 sec was 74.9±1.6 µM. to activate platelets, and the content of 5‑HT in supernatants Similar results were obtained when platelet aggregation was was measured using a Serotonin ELISA kit. As shown in induced by U46619 (1 µM) and AA (600 µM). LGP exhibited Fig. 4A, the level of serotonin was low when the platelets were INTERNATIONAL JOURNAL OF MOlecular medicine 42: 615-624, 2018 619

Figure 5. Effects of LGP on ADP‑ and U46619‑mediated TXA2 formation. (A) Effects of LGP on ADP‑induced TXA2 formation. (B) Effects of LGP Figure 4. Effects of LGP on ADP‑ and U46619‑mediated secretion of 5‑HT. on U46619‑induced TXA2 formation. As TXA2 is unstable and is rapidly (A) Effects of LGP on ADP‑induced 5‑HT release. (B) Effects of LGP on converted into a stable non‑enzymatic hydration product TXA2. A TXB2 U46619‑induced 5‑HT release. Each assay was performed in triplicate. content assay was performed. Each assay was performed in triplicate. **P<0.01, compared with control; ##P<0.01, compared with ADP or U46619. **P<0.01, compared with control; ##P<0.01, compared with ADP or U46619. Vehicle (DMSO) was used as a control, and ticagrelor and BM‑531 served Vehicle (DMSO) was used as a control, and ticagrelor and BM‑531 served as the positive controls. LGP, luteolin‑4'‑O‑β‑D‑glucopyranoside; ADP, as the positive controls. LGP, luteolin‑4'‑O‑β‑D‑glucopyranoside; ADP, adenosine diphosphate; 5‑HT, serotonin. adenosine diphosphate; TX, thromboxane.

treated with vehicle, whereas ADP (10 µM) treatment induced As shown in Fig. 6A and B, LGP significantly reduced the a sharp increase in the level of 5‑HT. The positive control expression of phospho‑RhoA in the presence of 1 µM U46619, (ticagrelor) significantly suppressed the increase induced by however, there was no significant change in the expression ADP. LGP also caused a significant reduction in the release of RhoA. Furthermore, it was found that the expression of serotonin in a concentration‑dependent manner, with the of ARHGEF1 (p115RhoGEF), an activator of RhoA, and inhibitory percentages of 22.55, 38.04, 47.71 and 65.95%, phospho‑ROCK1, an effecter of RhoA, were decreased by respectively. Similarly, LGP decreased the release of 5‑HT LGP in a dose‑dependent manner. stimulated by U46619 (1 µM) in a concentration‑dependent manner (Fig. 4B). Effects of LGP on regulating PI3K/Akt/GSK3β signal trans‑ duction stimulated by ADP. Subsequently, the present study

LGP inhibits ADP and U46619‑induced TXA2 synthesis. examined the effects of LGP on PI3K/Akt/GSK3β signal

TXA2, which is produced by activated platelets, serves to transduction stimulated by ADP. Western blot analysis with promote further platelet activation by binding to TP receptor. platelet lysates revealed that pre‑incubation of the platelets The present study examined the effects of LGP on the content with LGP (40, 80 and 160 µM) attenuated the expression ofTXB2 in platelets treated with ADP and U46619. As shown of phospo‑PI3K in the presence of ADP. Consistently, the in Fig. 5A, ADP markedly stimulated TXB2 release, whereas phosphorylation of Akt and GSK3β were also suppressed by LGP and ticagrelor caused a significant inhibition in the forma- LGP (40, 80 and 160 µM; Fig. 7A and B). tion of TXB2 in a dose‑dependent manner. Similar results were obtained with platelet activatorU46619 (Fig. 5B). Effects of LGP on platelet P2Y12 and TP receptor binding of [3H]‑2‑MeS‑ADP and [3H] SQ‑29548. In order to further Effects of LGP on RhoA signaling induced by U46619. One define whether the LGP‑mediated inhibitory effects on platelet previous study with inhibitors and/or genetic manipula- activation are due to antagonism at P2Y12 and TP receptors, tions has demonstrated that RhoA signaling contributes the present study performed a radiolabeled ligand binding 3 3 to TXA2‑induced platelet activation by binding the TP assay using [ H]‑2‑MeS‑ADP and [ H]SQ‑29548. LGP inhib- receptor (18). In order to determine whether TP‑mediated ited the binding of [3H]‑2‑MeS‑ADP to rat platelet membranes signal transduction is involved in the inhibitory effects of LGP with Ki=0.8317 mM as shown in Fig. 8A. LGP also exhibited on platelet activation, the present study examined the effects apparent competing effects on the TP receptor, which displaced of LGP on the activation of RhoA signaling transducers. [3H] SQ‑29548, a high affinity ligand of TP receptor from rat 620 XU et al: ANTIPLATELET EFFECTS OF LGP

Figure 6. Effects of LGP on U46619‑induced RhoA signaling. (A) Western blot analysis of protein lysates with (B) densitometric analysis of electrophoretic bands. Each assay was performed in triplicate. **P<0.01, compared with control; ##P<0.01, compared with U46619. Vehicle (DMSO) was used as a control, and BM‑531 served as the positivecontrol. LGP, luteolin‑4'‑O‑β‑D‑glucopyranoside; ADP, adenosine diphosphate; RhoA, Ras homolog family member A; ROCK1, Rho‑associated kinase 1; ARHGEF1, Rho, guanine nucleotide exchange factor 1; p‑, phosphorylated.

platelet membranes with Ki=1.520 mM (Fig. 8B). In addition, Discussion ticagrelor, a selective P2Y12 , and BM‑531, a selective TP receptor antagonist, displaced [3H]‑2‑MeS‑ADP Platelets have a major role in thromboembolic diseases, which and [3H]SQ‑29548 from their respective receptors at concen- are the final events complicating cardiovascular diseases trations in the nanomolar range (Fig. 8C and D). From these and peripheral vascular diseases (19). Therefore, antiplatelet data, the dose‑dependent displacement of [3H]‑2‑MeS‑ADP therapy remains crucial for patients with these diseases in and [3H] SQ‑29548 from their receptors by LGP was observed treatment and prophylaxis (19). However, the multiple pathways in rat platelets; however, compared with the effects of selective of platelet activation limit the effects of currently available antagonists of these receptors, the Ki values were higher than antiplatelet agents, resulting in limited clinical efficacy. The for the selective antagonists. These data indicated that LGP efficacy of existing antiplatelet therapies cannot be dissociated exhibited weak dual receptor inhibitory effects on P2Y12 and from an increased risk of bleeding (20,21). Previous lessons TP receptors. have demonstrated that, despite the implementation of existing INTERNATIONAL JOURNAL OF MOlecular medicine 42: 615-624, 2018 621

Figure 7. Effects of LGP on ADP‑induced PI3K/Akt/GSK3β signal transduction. (A) Western blot analysis of protein lysates with (B) densitometric analysis of electrophoretic bands. Western blot analysis was performed. Each assay was performed in triplicate. **P<0.01, compared with control; ##P<0.01, compared with ADP. Vehicle (DMSO) was used as a control, and ticagrelor served as the positive control. LGP, luteolin‑4'‑O‑β‑D‑glucopyranoside; ADP, adenosine diphosphate; PI3K, phosphoinositide 3‑kinase; GSKβ, glycogen synthase kinase 3β; p‑, phosphorylated. treatments, the incidence of side events remains high (20). Based on the screening results (data not shown), the initial Therefore, the development of effective and safe methods to experiments performed in the present study were designed to inhibit platelet function remains critically important. Several confirm the effects of LGP on platelet aggregation induced medicinal herbs exhibit antiplatelet effects, and these herbs by different agonists. It was found that LGP caused the are used in traditional Chinese medicine for promoting blood concentration‑dependent inhibition of the platelet aggregation circulation. In our previous study, the effects of several deriva- induced by ADP, U46619 and AA. In general, the formation of a tives of luteolin on ADP‑induced platelet aggregation were stable platelet plug occurs in three distinct steps: Platelet adhe- evaluated, and it was found that LGP exhibited significant sion, platelet activation and platelet aggregation (22). Platelet suppressive effects. The present study reported on the anti- adhesion to sites of vascular injury is triggered by exposure platelet effects of LGP, a flavonoid from C. nudiflora which is of the subendothelial extracellular matrix (ECM) following used as a treatment for promoting blood circulation in China. vascular injury, which is mediated by platelet interactions 622 XU et al: ANTIPLATELET EFFECTS OF LGP

3 3 Figure 8. Effects of LGP on platelet P2Y12 and TP receptor binding of [ H]‑2‑MeS‑ADP and [ H] SQ‑29548. Displacement curves show the specific binding induced by LGP. (A) [3H]‑2‑MeS‑ADP displacement curve of the specific binding induced by LGP. (B) [3H] SQ‑29548 displacement curve of the specific binding induced by LGP. (C) [3H]‑2‑MeS‑ADP displacement curve of the specific binding induced by ticagrelor. (D) [3H] SQ‑29548 displacement curve of the specific binding induced by BM‑531. LGP, luteolin‑4'‑O‑β‑D‑glucopyranoside; ADP, adenosine diphosphate; TP, thromboxane receptor.

with ECM components, particularly Von Willebrand factor, change, phosphoinositide hydrolysis, Ca2+ mobilization, collagen, fibronectin, thrombospondinand laminin. The adhe- protein phosphorylation and secretion, further amplifying the sion between platelets and ECM leads to the deceleration of activation signaling (24,25). It was also observed in the present flowing platelets and capture of circulating platelets to the study that LGP caused a significant decrease in the production vessel wall. At least two separate receptors on platelet cell of TXA2. LGP also inhibited the activation of αIIbβ3 integrin membranes, GpIb‑V‑IX complex and GpVI, function to tether in a dose‑dependent manner. The activation of integrin, the platelet and initiate cellular activation, simultaneously. particularly αIIbβ3 integrin, is considered the most important Once the platelets have adhered to the damaged vascular endo- step in platelet aggregation. Specific interactions of agonists thelium, they recruit additional platelets from the circulation with their receptors generate inside‑out signaling, leading to to augment the fragile platelet monolayer and eventually form the conformational activation of integrins, particularly αIIbβ3 a stable plug (23). Following platelet activation, a coordinated integrin, increasing their ligand affinity. The binding of αIIbβ3 series of events is triggered, including a rapid conformational integrin to its ligands, mainly fibrinogens, supports processes change and the secretion of α,δ‑granules and other intracel- including the close contact between aggregated platelets, lular vesicles, including lysosomes, which provide a positive and eventually promotes platelet activation and aggregation. feedback signal during platelet activation (19). Led by these The present study demonstrated that ADP, U46619 and AA concepts, the present study examined the effects of LGP on induced platelet aggregation, and α,δ‑granule release and the content of platelet granule components to further confirm TXA2 synthesis were inhibited by LGP in a dose‑dependent the effects of LGP on platelet activation. It was found that the manner. These data suggested that the inhibitory effects of secretion of 5‑HT, an agonist of platelet activation stored in LGP on aggregation may be associated with its suppression of

δ‑granules, was significantly decreased by LGP. TXA 2, a labile platelet activation. synthesized by activated platelets, is referred to as a ADP is a critical mediator of platelet activation. By second wave mediator of platelet activation (24). The synthesis binding to its receptor, this agonist leads to full activation 2+ of TXA2 is mediated by a cascade of , including events, including platelet conformational change, Ca influx, 2+ ‑1; this is activated by elevated Ca , TXA2 synthesis and granule secretion. Additionally, ADP which induces translocation to the plasma membrane and is released from the δ‑granules of activated platelets and phosphorylation by the stress kinase P38 and extracellular amplifies its own effects (26). Therefore, ADP, in addition to signal‑regulated kinase 1/2. Once synthesized, it diffuses TXA2 and 5‑HY, are termed second wave mediators, which across the platelet membrane and causes conformational are released from platelets and amplify effects of platelet INTERNATIONAL JOURNAL OF MOlecular medicine 42: 615-624, 2018 623 activation (26). There are two distinct G‑protein‑coupled Availability of data and materials ADP receptors expressed on the surface of human platelets,

P2Y1 and P2Y12. Several studies have suggested that P2Y12 is The analyzed data sets generated during the study are avail- the major receptor in the amplification of platelet activation, able from the corresponding author upon reasonable request. and the PI3K/PDK1/Akt/GSK3β pathway, particularly p110β and p110γ PI3K isoforms, has emerged as a major signaling Authors' contributions axis regulating P2Y12‑mediated platelet activation (27‑29). As mentioned above, TXA2 is another stimulator amplifying HX performed the platelet aggregation assay, western blot platelet activation, which is synthesized by activated platelets. analysis and the receptor‑binding assay. XZ contributed to

Once synthesized and diffused from platelets, TXA2 activates platelet preparation and integrain assay. WW was responsible and recruits platelets to the growing platelet aggregation via for 5‑HT and TXA2 measurement. ZZ participated in the the Gq‑ and G12/13‑coupled thromboxane-prostanoid receptors receptor‑binding assay. HF participated in the western blot TPα and TPβ (30). It is accepted that Rho GTPase signaling is analysis. SM and YL were responsible for LGP preparation involved in TP‑mediated platelet activation by causing platelet and wrote the manuscript. JF took responsibility for the design conformational change and regulating platelet secretion (31). of this project, analysis and interpretation of data. All authors Huang et al found that RhoA was activated by ARHGEF1 read, edited and approved the final manuscript. when platelets were stimulated by U46619, a mimetic of

TXA2 (32). Therefore, in order to determine whether the Ethics approval and consent to participate inhibitory effects of LGP on ADP‑induced platelet aggrega- tion were due to P2Y12‑mediated signaling inhibition, the The protocol used for animal experiments (JZAEC‑2016‑0031) effects of LGP on the activities of PI3K/PDK1/Akt/GSK3β was approved by the Animal Ethics Committees of Jiangxi were examined. LGP led to a dose‑dependent decrease in the University of Traditional Chinese Medicine, and all animal expression of p‑PI3K (Tyr607), p‑AktSer473, Thr308) and experiments were performed in strict accordance with the p‑GSK3β (Ser9) (Fig. 7). Similarly, it was found that LGP requirements of this protocol inhibited U46619‑induced RhoA signaling (Fig. 6). These data indicated that the signal transduction mediated by P2Y12 Consent for publication and TP receptors was involved in the LGP‑induced platelet inhibition. To further assess this possibility, the effects of LGP Not applicable on P2Y12 and TP receptors were evaluated by a radiolabeled ligand binding assay. As shown in Fig. 8, the dose‑dependent Competing interests displacement of [3H]‑2‑MeS‑ADP and [3H] SQ‑29548 from their receptors was caused by LGP in rat platelets, however, The authors declare that they have no competing interests. the Ki values were higher compared with the selective antago- nists. These data confirmed the effects of LGP on the activities References of P2Y12 and TP receptors and downstream signal transduc- tion. However, how LGP affects the P2Y12 and TP receptors 1. Harrison P: Platelet function analysis. Blood Rev 19: 111‑123, 2005. 2. Ghoshal K and Bhattacharyya M: Overview of platelet remains to be fully elucidated. Future investigations will focus physiology: Its hemostatic and nonhemostatic role in disease on the association between the chemical structure of LGP and pathogenesis. ScientificWorldJournal 2014: 781857, 2014. the P2Y and TP receptors, to elucidate why and how this 3. Weyrich AS: Platelets: More than a sack of glue. Hematology Am 12 Soc Hematol Educ Program 2014: 400‑403, 2014. compound affects P2Y12 and TP receptors. 4. Clemetson KJ: Platelets and primary haemostasis. Thromb In conclusion, the data presented in the present study Res 129: 220‑224, 2012. demonstrated that LGP, a natural compound from C. nudiflora 5. Yun SH, Sim EH, Goh RY, Park JI and Han JY: Platelet Activation: The mechanisms and potential biomarkers. Biomed Hook, inhibited the development of platelet aggregation and Res Int 2016: 9060143, 2016. amplification of platelet activation. These inhibitory effects 6. Bye AP, Unsworth AJ and Gibbins JM: Platelet signaling: A complex interplay between inhibitory and activatory networks. may be associated with its dual‑receptor inhibition on P2Y12 J Thromb Haemost 14: 918‑930, 2016. and TP receptors. 7. Michelson AD: Antiplatelet therapies for the treatment of cardio- vascular disease. Nat Rev Drug Discov 9: 154‑169, 2010. 8. Jackson SP and Schoenwaelder SM: Antiplatelet therapy: In search Acknowledgements of the ‘magic bullet’. Nat Rev Drug Discov 2: 775‑789, 2003. 9. Sigalov AB: Novel mechanistic concept of platelet inhibition. The authors would like to thank Dr Xianghua Xu and Expert Opin Ther Targets 12: 677‑692, 2008. 10. Varon D and Spectre G: Antiplatelet agents. Hematology Am Soc Dr Yaping Pan from Baylor College of Medicine (Houston, Hematol Educ Program: 267‑272, 2009. Texas) for critically reading the manuscript. 11. Pei J and Chen SL: Flora of China. Beijing: Science Press 1982. 12. Zhou Z, Wei X, Fu H and Luo Y: Chemical constituents of Callicarpa nudiflora and their anti‑platelet aggregation activity. Funding Fitoterapia 88: 91‑95, 2013. 13. Luo YH, Ma SC, Hu SR, Fu HZ, Zhou ZQ and Chen WK: Chemical constituents from callicarpa nudiflora. Zhong Yao The present study was supported by grants from the National Cai 38: 2306‑2310, 2015 (In Chinese). Natural Science Foundation of China (grant nos. 81373955, 14. Del Turco S, Sartini S, Cigni G, Sentieri C, Sbrana S, Battaglia D, 81560639 and 81660680) and the Natural Science Foundation Papa A, Da Settimo F, La Motta C and Basta G: Synthetic analogues of flavonoids with improved activity against platelet of Jiangxi Province (grant nos. 20142BAB205085 and activation and aggregation as novel prototypes of food supple- 20171BAB205097). ments. Food Chem 175: 494‑499, 2015. 624 XU et al: ANTIPLATELET EFFECTS OF LGP

15. Armstrong PC, Truss NJ, Ali FY, Dhanji AA, Vojnovic I, Zain ZN, 24. Nakahata N: Thromboxane A2: Physiology/pathophysiology, Bishop‑Bailey D, Paul‑Clark MJ, Tucker AT, Mitchell JA and cellular signal transduction and pharmacology. Pharmacol Warner TD: and the in vitro linear relationship between Ther 118: 18‑35, 2008. thromboxane A2‑mediated platelet aggregation and platelet 25. Stegner D and Nieswandt B: Platelet receptor signaling in production of thromboxane A2. J Thromb Haemost 6: 1933‑1943, thrombus formation. J Mol Med 89: 109‑121, 2011. 2008. 26. Goggs R and Poole AW: Platelet signaling‑a primer. J Vet Emerg 16. Schoenwaelder SM, Ono A, Sturgeon S, Chan SM, Mangin P, Crit Care 22: 5‑29, 2012. Maxwell MJ, Turnbull S, Mulchandani M, Anderson K, 27. Gurbel PA, Kuliopulos A and Tantry US: G‑protein‑coupled Kauffenstein G, et al: Identification of a unique co‑operative receptors signaling pathways in new develop- phosphoinositide 3‑kinase signaling mechanism regulating ment. Arterioscler Thromb Vasc Biol 35: 500‑512, 2015. integrin alpha IIb beta 3 adhesive function in platelets. J Biol 28. Von Kügelgen I and Hoffmann K: Pharmacology and structure Chem 282: 28648‑28658, 2007. of P2Y receptors. Neuropharmacology 104: 50‑61, 2016. 17. Gao J and Shattil SJ: An enzyme‑linked immunosorbent assay 29. Cattaneo M: P2Y12 receptors: Structure and function. J Thromb to identify inhibitors of activation of platelet integrin alpha IIb Haemost 13 (Suppl 1): S10‑S16, 2015. beta 3. J Immunol Methods 181: 55‑64, 1995. 30. Huang JS, Ramamurthy SK, Lin X and Le Breton GC: Cell 18. Gratacap MP, Payrastre B, Nieswandt B and Offermanns S: signalling through thromboxane A2 receptors. Cell Signal 16: Differential regulation of Rho and Rac through heterotrimeric 521‑533, 2004. G‑proteins and cyclic nucleotides. J Biol Chem 276: 47906‑47913, 31. Goggs R, Williams CM, Mellor H and Poole AW: Platelet 2001. 19. Angiolillo DJ, Ueno M and Goto S: Basic principles of platelet Rho GTPases‑a focus on novel players, roles and relationships. biology and clinical implications. Circ J 74: 597‑607, 2010. Biochem J 466: 431‑442, 2015. 20. Yousuf O and Bhatt DL: The evolution of antiplatelet therapy in 32. Huang JS, Dong L, Kozasa T and Le Breton GC: Signaling cardiovascular disease. Nat Rev Cardiol 8: 547‑559, 2011. through Gα13 switch region I is essential for protease‑activated 21. Gachet C: Antiplatelet drugs: Which targets for which treat- receptor 1‑mediated human platelet shape change, aggregation, ments? J Thromb Haemost 13 (Suppl 1): S313‑S322, 2015. and secretion. J Biol Chem 282: 10210‑10222, 2007. 22. De Meyer SF, Vanhoorelbeke K, Broos K, Salles II and Deckmyn H: Antiplatelet drugs. Br J Haematol 142: 515‑528, 2008. 23. Shifrin MM and Widmar SB: Platelet inhibitors. Nurs Clin North Am 51: 29‑43, 2016.